[cairo] Proper invocation of cairo_clip()?

[Bob Jamison]

Owen Taylor wrote:

>On Tue, 2005-03-15 at 13:50 -0600, Bob Jamison wrote:
>  
>
>>Thanks for the reply!
>>
>>Owen Taylor wrote:
>>
>>
>>    
>>
>>>Well, it shouldn't segfault ... so file a bug in bugzilla.
>>>      
>>>
>>Never mind.  The width param was 0.0, resulting in no traps.
>> (Maybe the function should check this?)
>>    
>>
>
>If you get cairo to segfault, file a bug in bugzilla.
>
>  
>
>>>But it probably doesn't what you want... (not exactly
>>>sure what that is, but that call establishes a clip 
>>>of [x,y,width,height] in device space, so after that call
>>>"do my drawing" is going to have to draw at 2*x, 2*y to
>>>not be clipped.)
>>>
>>>      
>>>
>>Here is what I am trying to do: take a subrect of the drawing
>>surface and use it in vacuo of the rest of the drawing, like
>>a virtual drawing surface:
>>
>>http://troi.hous.es3.titan.com/inkscape/files/clip-n-translate.png
>>
>>Am I trying to do something out of bounds?  Maybe cairo_copy()
>>has a purpose?
>>    
>>
>
>cairo_copy() certainly isn't related here. If you just want to
>draw on a subportion of your drawable (x,y,width,height) with
>
> clipping to that rectangle 
> (0,0) at the upper left of that rectangle
>
>Then your code was basically right other than the sign of 
>cairo_translate() being backwards. (Think of cairo_translate
>as moving the axes)
>
>Regards,
>						Owen
>
>  
>
I filed the bug (if it's a bug).

And removing the '-' worked great.
I had assumed it was shifted the other way, from the old days
of using OpenGL.  Bad habits.

All is working fine now.  Thanks!


Bob